Semiconductor device and power conversion apparatus using the same

Abstract

 A semiconductor device provides a gate electrode (401) formed on a lateral face of a wide trench (423), and thereby cover the gate electrode is covered by a gate insulating layer (402) and a thick insulating layer (403) to be an inter layer. Therefore, a parasitic capacitance of the gate becomes small, and there is no potential variation of the gate since there is no floating p-layer so that a controllability of the dv/dt can be improved. In addition, the conductive layer between the gate electrodes can relax the electric field applied to the corner of the gate electrode. In consequence, compatibility of low loss and low noise can be achieved.

Description

BACKGROUND OF THE INVENTION

[0001] The present invention relates to a semiconductor device and a power conversion apparatus using the same, and in particular to a semiconductor device preferred for an insulated gate bipolar transistor (hereinafter abbreviated to IGBT) having an insulated trench gate structure and a power conversion apparatus using the same.

[0002] The IGBT is a switching device for controlling a current to be flown between a collector electrode and an emitter electrode in response to a voltage applied to a gate electrode. Since electric power to be controlled by IGBT is ranged from several tens to several hundred thousand watts and switching frequency is ranging from several tens to over hundred kilohertz, the IGBT has been widely used from low electric power devices, such as an air-conditioner, microwave oven, etc. for household use to high electric power devices, such as an inverter for use in railroad rolling stock, steel plant, etc.

[0003] For the purpose of realizing a high efficiency operation of the electric devices, IGBT has been demanded for a low loss, that is, demanded for reduction of a conduction loss and a switching loss. At the same time, it has been demanded that dv/dt can be controlled to fit a specification for an application so as to prevent a problem, such as EMC noise, malfunction, breakdown of motors, etc.

[0004] Incidentally, JP-A-2000-307116 discloses IGBT having a structure in which an arrangement pitch of a trench gate is changed, as illustrated in FIG 23. A feature of the IGBT in FIG 23 is that a floating p-layer 105 is formed on a portion where the interval between the trench gates is wide, without forming a p-channel layer 106.

[0005] In the configuration as mentioned above, since a current is only flown into a portion where the interval between the trench gates is narrower, an over current flowing at a short circuit can be inhibited, therefore, a ruggedness of the device can be improved. Further, since a part of a hole current flows into an emitter electrode 114 via the floating p-layer 105, a hole concentration increases around the emitter electrode 114, therefore, there is also an advantage that an on-state voltage can be reduced. Furthermore, there is also an advantage that an electric field applied to the trench gate is relaxed by a p-n junction formed of the floating p-layer 105 and an n^- drift layer 104.

SUMMARY OF THE INVENTION

[0006] However, in the IGBT illustrated in FIG 23, a problem sometimes occurs that controllability of the time change of an output voltage, dv/dt, of the IGBT and opposite-arm diode reduces when the IGBT turns on. FIG 24 represents a simulated waveforms of the collector-emitter voltage vce during a turn-on operation. As illustrated in FIG 24, there is a time period during which the control cannot be performed, that is, dvce/dt does not change even though a gate resistance is changed.

[0007] This reason is considered, as follows. When the IGBT turnes on, holes flow transiently into the floating p-layer 105 in FIG 23, and an electric potential at the floating p-layer 105 becomes high. At this time, since a displacement current flows into a gate electrode 109 via a parasitic capacitance of a gate insulating layer 110 and thereby a gate voltage is raised, a time change of a collector current, dic/dt, determined by the product of a mutual conductance, gm, in a MOSFET structure and a time change of a voltage between the gate and emitter, dvge/dt, increases, and thereby, a switching speed is accelerated. The amount of holes transiently flown into the floating p-layer 105 is determined mainly by the structure inside the semiconductor. Therefore, it is difficult to control the amount of holes by an outside gate resistance.

[0008] Therefore, the accelerated dic/dt cannot be controlled by the outside gate resistance. As a result, the time period occurs during which dv/dt of the IGBT and the opposite-arm diode cannot be controlled by the gate resistance.

[0009] For the purpose of inhibiting the gate potential rise caused by the effect of the floating p-layer 105, the following methods have been proposed.

[0010] In JP-A-2004-39838, as illustrated in FIG 25, the floating p-layer 105 is electrically connected to the emitter electrode 114 via a resistor 201, therefore, the electric potential rise at the floating p-layer 105 is inhibited.

[0011] In this way, the displacement current flowing into the gate electrode 109 from the floating p-layer 105 decreases, and therefore, inhibits the gate potential rise. As a result of this, the controllability of the dv/dt of the IGBT and the opposite-arm diode can be improved.

[0012] JP-A-5-243561 discloses that a wide trench gate is formed, as illustrated in FIG 26, to enhance an injection of electrons from the trench gate, and therefore, the loss of the IGBT is reduced. Although JP-A-5-243561 does not explicitly specify, because of no floating p-layer, there is no gate potential variation caused by the effect of the floating p-layer, and therefore, the controllability of the dv/dt can be improved.

[0013] JP-A-2009-200103 discloses that a gate electrode 301 is formed on a lateral face of the trench and a space between the gate electrodes 301 is filled up with an insulating layer 302, and thereby the parasitic capacitance of a gate insulating layer 110 is reduced, as illustrated in FIG 27. Therefore, the gate voltage rise can be controlled and the controllability of the dv/dt can be improved.

[0014] A document "Y. Onozawa, et al., Proc. 19th ISPSD, pp13-16, 2007" discloses that, as illustrated in FIG 28, a p base layer 106 electrically connected with the emitter electrode is alternately disposed toward a direction in parallel with the trench at every cell without using the floating p-layer. Therefore, there is no gate potential variation caused by the effect of the floating p-layer and the controllability of the dv/dt can be improved because the floating p-layer is not used.

[0015] Incidentally, it has been demanded that the dv/dt controllability by the gate drive circuit during a turn-on switching time period is improved while maintaining the low loss and high breakdown voltage in IGBT. For this problem, it was appreciated that the structure in the documents as mentioned above has the following improvement points.

[0016] In the JP-A-2004-39838, the smaller the resistance value of the resistor 201 between the floating p layer 105 and emitter electrode 114 becomes, the more the controllability of the dv/dt is improved. However, since a part of the hole current at on-state flows out to the emitter electrode 114 via the resistor 201, the effect of enhancing the injection of electrons becomes small, and thereby, the on-voltage is increased. Conversely, when increasing the resistance value of resistor 201, the on-voltage rise becomes small, while there is a problem that the controllability of the dv/dt is reduced.

[0017] In the JP-A-5-243561, there is a problem that, since the wide trench gate is formed and the parasitic capacitance of the gate is therefore large, a reverse transfer capacitance of IGBT becomes large, and thereby, a switching loss of IGBT and an electric power of the gate drive circuit is increased.

[0018] In the JP-A-2009-200103, there is a problem that the breakdown voltage and reliability become reduced, because the width of trench is made large to enhance the injection of electrons, and thereby, the electric field of the bottom of a gate electrode 301 formed on a lateral face is increased.

[0019] In the document "Y Onozawa, et al., Proc. 19the ISPSD, pp13-16 2007", there is a problem that the electric field applied to the gate electrode 109 increases and thereby the breakdown voltage and reliability is reduced, because the p base layers 106 are partially removed.

[0020] The present invention is made in light of the above-mentioned points. An object of the invention is to provide a semiconductor device capable of improving the dv/dt controllability by the gate drive circuit during the turn-on switching time period, while maintaining the low loss and high breakdown voltage, and to provide a power conversion apparatus using the same.

[0021] In order to achieve the above-mentioned object, a semiconductor device of the invention comprises: a first semiconductor layer of a first conductive type; a second semiconductor layer of a second conductive type formed in the vicinity of a surface of the first semiconductor layer; a third semiconductor layer of the second conductive type adjacent to the first semiconductor layer and formed in the vicinity of a surface of an opposite side of the second semiconductor layer; a fourth semiconductor layer of the first conductive type selectively formed on the third semiconductor layer; a trench penetrating through the fourth semiconductor layer and the third semiconductor layer into the first semiconductor layer; a gate insulating layer formed along an internal face of the trench; an insulating layer formed in the trench; a first conductive layer formed on a surface of the gate insulating layer; and a second conductive layer formed on a surface of the insulating layer, characterized in that the first conductive layer in the trench of at least one cross-sectional structure is divided by the insulating layer and the second conductive layer, a width "a" of the trench is wider than a width "b" of an area in which the trench is not formed, and the first conductive layer and the gate insulating layer are formed on a lateral face of the trench.

[0022] Further, a power conversion apparatus of the invention comprises: a pair of input terminals; a plurality of series connected circuits, connected between the input terminals, to which a plurality of semiconductor switching devices are connected in series; a plurality of output terminals each connected to a series connected point of each of the plurality of series connected circuits, and the plurality of semiconductor switching devices that turn on and off to perform a power conversion, characterized in that the plurality of semiconductor switching devices are above-mentioned semiconductor devices.

[0023] The semiconductor device and the power conversion apparatus using the same in the invention have an effect that the reverse transfer capacitance can be reduced, and thereby, the dv/dt controllability by the gate drive circuit during the turn-on switching time period can be improved, while maintaining the low loss and high breakdown voltage.

[0024] The other objects, aspects and advantages of the invention will become apparent from the following description of the embodiments of the invention taken in conjunction with the accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

[0025] 

FIG 1 is a cross-section diagram illustrating an IGBT as a semiconductor device according to a first embodiment of the invention;

FIG 2 is a simulated characteristic diagram representing a collector voltage dependency of the reverse transfer capacitances of the IGBT according to the first embodiment of the semiconductor of the invention and in the JP-A-2000-307116;

FIG 3 is a characteristic diagram illustrating the simulated waveforms of collector

• emitter voltage during the turn-on operation of an IGBT according to the first embodiment of the semiconductor of the invention;

FIG 4 is a diagram illustrating a manufacturing process of an IGBT according to the first embodiment of the semiconductor of the invention;

FIG 5 is a cross-section diagram illustrating an IGBT according to a second embodiment of the semiconductor of the invention;

FIG 6 is a cross-section diagram of a termination area of an IGBT according to the second embodiment;

FIG 7 is a cross-section diagram illustrating an IGBT according to a third embodiment of the semiconductor of the invention;

FIG 8 is a cross-section diagram illustrating an IGBT according to a fourth embodiment of the semiconductor of the invention;

FIG 9 is a cross-section diagram illustrating an IGBT according to an example of the fourth embodiment of the semiconductor of the invention;

FIG 10 is a cross-section diagram illustrating an IGBT according to an other example of the fourth embodiment of the semiconductor of the invention;

FIG 11 is a cross-section diagram illustrating an IGBT according to a fifth embodiment of the semiconductor of the invention;

FIG 12 is a cross-section diagram illustrating an IGBT according to an example of the fifth embodiment of the semiconductor of the invention;

FIG 13 is a cross-section diagram illustrating an IGBT according to an other example of the fifth embodiment of the semiconductor of the invention;

FIG. 14 is a cross-section diagram illustrating an IGBT according to a sixth embodiment of the semiconductor of the invention;

FIG 15 is a cross-section diagram illustrating an IGBT according to a seventh embodiment of the semiconductor of the invention;

FIG 16 is a cross-section diagram illustrating an IGBT according to an example of the seventh embodiment of the semiconductor of the invention;

FIG 17 is a cross-section diagram illustrating an IGBT according to an other example of the seventh embodiment of the semiconductor of the invention;

FIG 18 is a cross-section diagram illustrating an IGBT according to an other example of the seventh embodiment of the semiconductor of the invention;

FIG 19 is a cross-section diagram illustrating an IGBT according to an other example of the seventh embodiment of the semiconductor of the invention;

FIG 20 is a cross-section diagram illustrating an IGBT according to an eighth embodiment of the semiconductor of the invention;

FIG 21 is a cross-section diagram illustrating an IGBT according to a ninth embodiment of the semiconductor of the invention;

FIG 22 is a circuit diagram representing one embodiment of a power conversion apparatus of the invention;

FIG 23 is a cross-section diagram illustrating a conventional IGBT disclosed in the JP-A-2000-307116;

FIG 24 is a characteristic diagram illustrating a simulated waveforms of the collector- emitter voltage during the turn-on operation of the conventional IGBT disclosed in the JP-A-2000-307116;

FIG 25 is a cross-section diagram illustrating the conventional IGBT disclosed in the JP-A-2004-39838;

FIG 26 is a cross-section diagram illustrating the conventional IGBT disclosed in the JP-A-5-243561;

FIG 27 is a cross-section diagram illustrating the conventional IGBT disclosed in the JP-A-2009-200103;

FIG. 28 is a perspective diagram illustrating a partly broken portion of the conventional IGBT disclosed in the document "Y Onozawa, et al., Proc. 19th ISPSD, pp13-16 2007"; and

FIG 29 is a characteristic diagram for explaining grounds of a thickness of the insulating layer in the second embodiment.

DETAILED DESCRIPTION OF THE EMBODIMENTS

[0026] Hereinafter, a semiconductor device will be described in detail with reference to illustrated embodiments.

(First Embodiment)

[0027] FIG 1 shows a cross-sectional structure of an IGBT in a first embodiment of the semiconductor device of the invention.

[0028] The IGBT illustrated in FIG 1 is configured by: an n^- drift layer 104 as a first semiconductor layer of a first conductive type; a p-collector layer 102 as a second semiconductor layer of a second conductive type formed in the vicinity of the surface of n^- drift layer 104, and the layer 102 is contacted to a collector electrode 100; a p-channel layer 106 as a third semiconductor layer of the second conductive type adjacent to the n^- drift layer 104 and formed in the vicinity of the surface of a reverse side of the layer 104 with respect to the side of p-collector layer 102; an n-emitter layer 107 as a fourth semiconductor layer of the first conductive type formed selectively on the p-channel layer 106; a trench 423 penetrating the n-emitter layer 107 and the p-channel layer 106 into the n^- drift layer 104; a gate insulating layer 402 formed along an internal face of the trench 423; an insulating layer 403 formed in the trench 423; a gate electrode 401 as a first conductive layer formed on the surface of the gate insulating layer 402; an emitter electrode 404 as the second conductive layer formed on the surface of the insulating layer 403 and protruding partly in the trench 423 to the side of the n^- drift layer 104; an n-buffer layer 103 formed between the p-collector layer 102 and the n^- drift layer 104; and a p-contact layer 108 formed in the p-channel layer 106. In addition, a reference numeral 101 denotes a collector terminal, 115 denotes a gate terminal, and 116 denotes an emitter terminal.

[0029] The IGBT of this embodiment has the cross-sectional structure in which the gate electrode 401 is divided by the insulating layer 403 and the emitter electrode 404 in the trench 423, as illustrated in FIG 1. Further, a width "a" of the trench 423 is wider than a width "b" of an area in which the trench is not formed. A relation, a > b, is satisfied in FIG 1, and the gate electrode 401 is formed on the lateral face of wide trench 423. Since the gate electrode 401 is covered by the gate insulating layer 402 and the thick insulating layer 403 (for example, the gate insulating layer 402 is as thick as 100 nm, and the insulating layer (inter layer) 403 is as thick as 1000 nm), a parasitic capacitance of the gate can be drastically reduced, compared with the conventional IGBT illustrated in FIG 23. In the conventional IGBT, both sides of the gate are covered by a thin gate insulating layer. In contrast, in the structure of this embodiment, one side is covered by the gate insulating layer 402 and the other side is covered by the thick insulating layer 403, so that the parasitic capacitance of the thick inter layer side is reduced. A displacement current is given by (parasitic capacitance) x (potential variation of floating p-layer), that is, the smaller the parasitic capacitance, the smaller the displacement current. Therefore, the controllability of the dv/dt can be improved,because the potential rise in the gate is controllable due to the reduced displacement current.

[0030] FIG 2 represents a simulated result of a collector voltage dependency of a reverse transfer capacitance for the IGBT in this embodiment of the invention and the conventional IGBT. The parasitic capacitance of the gate is a reverse transfer capacitance component of the IGBT, so that the reverse transfer capacitance is reduced if the parasitic capacitance is reduced. As is apparent from FIG 2, in the IGBT of the first embodiment of the invention, the reverse transfer capacitance is reduced up to approximately 1/4 in comparison with the conventional IGBT.

[0031] Further, the IGBT in the first embodiment of the invention does not form the floating p-layer which has been formed in the conventional IGBT, therefore, there is no gate potential variation caused by an effect of the floating p-layer and the controllability of the dv/dt can be improved. A reduction of the breakdown voltage and the reliability is concerned about because the floating p-layer is not formed. However, the emitter electrode 404 is formed to protrude in the trench to the n^- drift layer 104 (the gate electrode 401 is divided by the emitter electrode 404 and the insulating layer 403), therefore, the reduction of the breakdown voltage and reliability can be prevented, because the emitter electrode 404 acts as a field plate which relaxes the electric field applied to a corner portion of the gate electrode 401.

[0032] FIG 3 illustrates simulated waveforms of the collector - emitter voltage, vce, on the turn-on operation of the IGBT according to the first embodiment of the invention. As is apparent from FIG 3, in the IGBT of the first embodiment of the invention, it is appreciated that dvce/dt can be controlled by varying a gate resistance, unlike the case of the conventional IGBT illustrated in FIG 24.

[0033] In FIG 4, (a) to (k) illustrate an example of a manufacturing process of the IGBT according to the first embodiment of the invention. First, a pattern as illustrated in (b) is formed on the n^- drift layer 104 illustrated in (a) using a photoresist 501, and the wide trench 423 is formed by an anisotropic etching as illustrated in (c). Next, the gate insulating layer 402 is formed as illustrated in (d), and polysilicon as a material of the gate electrode 401 is deposited, as illustrated in (e), to then be etched back by the anisotropic etching and form the gate electrode 401 made up of the polysilicon on the lateral face of the trench 423 as a sidewall, as illustrated in (f). Next, a pattern is formed using a photoresist 502 as illustrated in (g), and an ion implantation of p-type and n-type is performed to form the p-channel layer 106 and n-emitter layer 107 as illustrated in (h). Subsequently, the insulating layer 403 is deposited as illustrated in (i), a contact hole formed and the p-contact layer 108 is formed as illustrated in (j). Finally, the emitter electrode 404, the n-buffer layer 103, the p-collector layer 102 and the collector electrode 100 are formed, as illustrated in (k), and thus the process is completed.

[0034] In this embodiment, the collector terminal 101 and the p-collector layer 102 on the reverse face are formed after the surface process, however, an epitaxial-substrate on which the p-collector layer 102 are initially formed may be used.

[0035] As mentioned above, the gate electrode 401 is formed on the lateral face of the wide trench 423 in the IGBT in the first embodiment of the invention, so that the parasitic capacitance of the gate is reduced, and the dv/dt controllability by the gate drive circuit during the turn-on switching time period can be improved.

(Second Embodiment)

[0036] FIG 5 illustrates a cross-sectional structure of an IGBT according to a second embodiment of the invention. A feature of the IGBT in the second embodiment is that an insulating layer 403 is thinner than the thickness of the insulating layer 407 (refer to FIG 6) in a termination area (not shown, peripheral area of chip), and thereby enhance the field plate effect caused by the emitter electrode 404 and a higher breakdown voltage can be realized. As illustrated in FIG 5, as the corner portion (corner portion close to the emitter electrode 404 and on the lower side of the gate electrode 401) of the gate electrode 401 and an end portion (corner portion mostly close to the gate electrode 401) of the emitter electrode 404 become closer with each other, the field plate effect becomes higher. In consequence, the high breakdown voltage of the IGBT can be obtained.

[0037] However, in the termination area of the IGBT made up of a p-layer 405 and a floating electrode 406 as illustrated in FIG 6, if the insulating layer 407 becomes thin (similar extent to the insulating layer 403 as illustrated in FIG 5), the electric field in the insulating layer 407 becomes high and thereby reduces the breakdown voltage in the termination area. In the second embodiment, the insulating layer 403 in an active area (area in FIG 5) of the IGBT is thinner than the thickness of the insulating layer 407 in the termination area. In consequence, there is an advantage that both the active area and the termination area can maintain the high breakdown voltage. That is, the breakdown voltage of the active area is improved by using the thin insulating layer 403, while the breakdown voltage of the termination area is improved by using the thick insulating layer 407. Therefore, the feature of this embodiment is that the thickness of the insulating layer in the two areas is changed.

[0038] Specifically, it is desirable that the thickness of the insulating layer 403 ranges from 300 nm to 1000 nm and the thickness of insulating layer 407 is equal to or greater than 1000 nm. As illustrated in FIG 29, the breakdown voltage continues to improve in the range from 300 nm to 1000 nm in thickness of the oxide layer (insulating layer 403) and is saturated at an equal to or less than 300 nm in thickness. Further, if the oxide layer is too thin, impurities diffuse up to an Si-layer under the insulating layer when implanting ions. Therefore, the extent of around 300 nm is desirable.

(Third Embodiment)

[0039] FIG 7 illustrates a cross-sectional structure of an IGBT according to a third embodiment of the invention. A feature of the IGBT in the third embodiment is that a trench 424 has a two step staircase pattern, and the n^- drift layer 104 side of the emitter electrode 404 protrudes deeper than the corner portion of the gate electrode 401.

[0040] By forming the structure as in the third embodiment as mentioned above, the electric field applied to the corner portion of the gate electrode 401 is dispersed by the n^- drift layer 104 side of the emitter electrode 404 and the corner portion of the trench 424, so that a higher breakdown voltage can be obtained from the same reason as described in the second embodiment.

(Fourth Embodiment)

[0041] FIG 8 to FIG 10 illustrate a cross-sectional structure of an IGBT according to a fourth embodiment of the invention. A feature of the IGBT in the fourth embodiment is that a p-layer is inserted underneath the trench 423 to thereby allow a depletion layer to extend from the p-layer. Therefore, the electric fields at the corner portion of the gate electrode 401 and a trench corner portion are relaxed, so that a higher breakdown voltage can be obtained. In FIG 8, a p-layer 408 is partly inserted underneath trench 423. In FIG 9, a p-layer 409 is inserted so as to cover the corner portion of the gate electrode 401. In FIG 10, an n-layer 410 is inserted into the p-layer 409. The n-layer 410 in FIG 10 prevents holes from flowing into the floating p-layer 409 during the turn-on operation. In fact, since less holes flow into the n-layer by forming the n-layer inside the p-layer, the holes to flow into the floating layer are reduced. In consequence, there is an advantage that the potential rise in the floating p-layer is inhibited.

(Fifth Embodiment)

[0042] FIG 11 to FIG 13 illustrate a cross-sectional structure of an IGBT according to a fifth embodiment of the invention. A feature of the IGBT in the fifth embodiment is that a p-layer 418 is formed underneath the p-channel layer 106. In consequence, the electric field at the corner of gate electrode 401 is relaxed, so that a higher breakdown voltage can be obtained. In FIG 11, it is configured that the p-layer 418 is inserted underneath the p-channel layer. In FIG 12, a p-layer 419 is inserted so as to cover the corner portion of the gate electrodes 401. In FIG 13, a p-layers 420 are inserted so as to cover each one of the corner of the gate electrode 401.

(Sixth Embodiment)

[0043] FIG 14 is a cross-sectional structure of an IGBT according to a sixth embodiment of the invention. A feature of the IGBT in the sixth embodiment is that a capacitor made up of an insulating layer 426 and a polysilicon electrode 411 is formed on the trench 423. The polysilicon electrode 411 is connected to the emitter electrode 404. Since a part of hole current flowing underneath the trench 423 during the turn-on is used for charging up the capacitor, the potential rise underneath the trench 423 is inhibited. In consequence, there is an advantage that the gate potential rise can be reduced. Since the polysilicon electrode 411 acts as the field plate to relax the electric field at the corner portion of the gate electrode 401, there is also the advantage that the high breakdown voltage can be obtained.

(Seventh Embodiment)

[0044] FIG 15 to FIG 19 illustrate a cross-sectional structure of an IGBT according to a seventh embodiment of the invention. In FIG 15, a polysilicon electrode 412 is formed on the lateral face of a trench 425 likewise to the gate electrode 401, and the capacitor made up of the gate insulating layer 402 and the polysilicon electrode 412 is connected to the emitter electrode 404. In consequence, there is an advantage that the gate potential rise can be reduced during the turn-on operation. In FIG 16, a p-layer 413 is formed between the polysilicon electrodes 412 to relax the electric field applied to the corner portion of the polysilicon electrode 412, therefore, the high breakdown voltage can be obtained. In FIG 17, an n-layer 414 is formed on the p-layer 413 to prevent the holes from flowing into the floating p-layer 413 during the turn-on operation. In consequence, there is an advantage that the potential rise of the floating p-layer is inhibited. In FIG 18, a p-layer 415 is formed so as to cover the corner portion of polysilicon electrode 412, so that a high breakdown voltage can be obtained. Further, in FIG 19, an n-layer 416 is formed on the p-layer 415 to prevent the holes from flowing into the floating p-type 415 during the turn-on operation, and there is the advantage that the potential rise of the floating p-layer is inhibited.

(Eighth Embodiment)

[0045] FIG 20 illustrates a cross-sectional structure of an IGBT according to an eighth embodiment of the invention. A feature of the eighth embodiment is that an n-layer 421 is inserted underneath the p-channel layer 106. Since the n-layer 421 works as a barrier for the holes flowing into the emitter electrode 404, a hole concentration in the vicinity of the emitter increases. In consequence, the on-voltage can be further reduced.

(Ninth Embodiment)

[0046] FIG 21 illustrates a cross-sectional structure of an IGBT according to a ninth embodiment of the invention. A feature of the ninth embodiment is that a p-layer 422 is inserted underneath the n-layer 421. In the structure of the eighth embodiment, the higher the carrier concentration of the n-layer 421, the higher the barrier for the holes, so that the reduction effect of the on-voltage becomes high. However, the electric field intensity becomes strong at the n-layer 421 during off-state, and it is therefore expected that the breakdown voltage is reduced.

[0047] In this embodiment as mentioned above, the p-layer 422 is added to underneath the n-layer 421 to relax the electric field intensity at the n-layer 421. Therefore, the breakdown voltage can be maintained even if the carrier concentration is made high, so that the on-voltage can be further reduced.

(Tenth Embodiment)

[0048] FIG 22 is a circuit diagram representing a power conversion apparatus employing the IGBT described in the above-mentioned embodiments.

[0049] FIG 22 represents an inverter circuit diagram in the embodiments. A reference numeral 601 denotes a gate drive circuit, 602 denotes an IGBT, 603 denotes a diode, 604, 605 denote an input terminal, and 606 to 608 denote an output terminal. A power conversion apparatus is configured in which the IGBT described in the first to ninth embodiments is applied to the inverter circuit.

[0050] The IGBT in the embodiments mentioned above is applied to the power conversion apparatus to thereby realize a low loss and high reliability of the power conversion apparatus.

[0051] The same advantages as in the inverters described above can also be obtained in a power conversion apparatus such as a converter, chopper, etc.

[0052] It should be further understood by those skilled in the art that although the foregoing description has been made on embodiments of the invention, the invention is not limited thereto and various changes and modifications may be made without departing from the spirit of the invention and the scope of the appended claims.

Claims

1. A semiconductor device comprising:

a first semiconductor layer (104) of a first conductive type;

a second semiconductor layer (102) of a second conductive type formed in the vicinity of a surface of the first semiconductor layer (104);

a third semiconductor layer (106) of the second conductive type adjacent to the first semiconductor layer (104) and formed in the vicinity of a surface of an opposite side of the second semiconductor layer (102);

a fourth semiconductor layer (107) of the first conductive type selectively formed on the third semiconductor layer (106);

a trench (423; 424; 425) penetrating through the fourth semiconductor layer (107) and the third semiconductor layer (106) into the first semiconductor layer (104);

a gate insulating layer (402) formed along an internal face of the trench (423; 424; 425);

an insulating layer (403; 426) formed in the trench (423; 424; 425);

a first conductive layer (401) formed on a surface of the gate insulating layer (402); and

a second conductive layer (404) formed on a surface of the insulating layer (403; 426),

characterised in that the first conductive layer (401) in the trench (423; 424; 425) of at least one cross-sectional structure is divided by the insulating layer (403; 426) and the second conductive layer (404).

2. A semiconductor device comprising:

a first semiconductor layer (104) of a first conductive type;

a second semiconductor layer (102) of a second conductive type formed in the vicinity of a surface of the first semiconductor layer (104);

a third semiconductor layer (106) of the second conductive type adjacent to the first semiconductor layer (104) and formed in the vicinity of a surface of an opposite side of the second semiconductor layer (102);

a fourth semiconductor layer (107) of the first conductive type selectively formed on the third semiconductor layer (106);

a trench (423; 424; 425) penetrating through the fourth semiconductor layer (107) and the third semiconductor layer (106) into the first semiconductor layer (104);

a gate insulating layer (402) formed along an internal face of the trench (423; 424; 425);

an insulating layer (403; 426) formed in the trench (423; 424; 425);

a first conductive layer (401) formed on a surface of the gate insulating layer (402); and

a second conductive layer (404) formed on a surface of the insulating layer (403; 426),

characterised in that at least one portion of the second conductive layer (404) protrudes in the trench (423; 424; 425) to a side of the first semiconductor layer (104).

3. The semiconductor device according to claim 1 or 2, wherein
a width "a" of the trench (423) is wider than a width "b" of an area in which the trench (423) is not formed, and
the first conductive layer (401) and the gate insulating layer (402) are preferably formed on a lateral face of the trench (423).

4. The semiconductor device according to any preceding claim, wherein
the first conductive layer (401) is covered by the gate insulating layer (402) and the insulating layer (403), and
the insulating layer (403) is preferably thicker than the gate insulating layer (402).

5. The semiconductor device according to any preceding claim, wherein
the insulating layer (403) of an active area in the semiconductor device is thinner than the insulating layer (407) of a termination area in the semiconductor device, and
a thickness of the insulating layer (403) in the active area of the semiconductor device preferably ranges from 300 nm to 1000 nm.

6. The semiconductor device according to any preceding claim, wherein
the trench (424) is formed of a two step staircase pattern, and
the second conductive layer (404) in the trench (424) formed of the two step staircase pattern preferably protrudes deeper than the edge portion of the first conductive layer (401) facing the first semiconductor layer (104).

7. The semiconductor device according to any preceding claim, wherein
a fifth semiconductor layer (408; 409) of the second conductive type is formed on a first semiconductor layer side of the trench (423),
the fifth semiconductor layer (409) is preferably formed so as to cover a first semiconductor layer side of the first conductive layer (401), and
a sixth semiconductor layer (410) of the first conductive type is preferably formed in the fifth semiconductor layer (409).

8. The semiconductor device according to any preceding claim, wherein a seventh semiconductor layer (418; 419; 420) of the second conductive type is formed on a first semiconductor layer side of the third semiconductor layer (106).

9. The semiconductor device according to claim 8, wherein
the seventh semiconductor layer (419) covers at least one part of the first semiconductor layer side of the first conductive layer (401), or
the seventh semiconductor layer is divided into two layers (420), and the respective seventh semiconductor layers (420) cover at least one part of the first semiconductor layer side of the first conductive layer (401).

10. The semiconductor device according to any preceding claim, wherein a capacitor made up of the insulating layer (426) and a third conductive layer (411; 412) is formed in the trench (423).

11. The semiconductor device according to any preceding claim, wherein the third conductive layer (412) is formed on a part of a lateral face of the trench (425), and a capacitor made up of the third conductive layer (412) and the gate insulating layer (402) is electrically connected to the second conductive layer (404).

12. The semiconductor device according to claim 11, wherein
an eighth semiconductor layer (413; 415) of the second conductive type is formed between the third conductive layers (412), and
a ninth semiconductor layer (414; 416) of the first conductive type is preferably formed on an opposite side of the first semiconductor layer side of the eighth semiconductor layer (413).

13. The semiconductor device according to claim 12, wherein
the eighth semiconductor layer (415) is formed so as to cover the first semiconductor layer side of the third conductive layer (412), and
the ninth semiconductor layer (416) preferably protrudes deeper than an edge portion of the trench (425).

14. The semiconductor device according to any preceding claim, wherein
a tenth semiconductor layer (421) of the first conductive type is formed on a first semiconductor layer side of the third semiconductor layer (106), and
an eleventh semiconductor layer (422) of the second conductive type is preferably formed on a first semiconductor layer side of the tenth semiconductor layer (421).

15. A power conversion apparatus comprising:

a pair of input terminals (604, 605);

a plurality of series connected circuits (601, 602, 603), connected between the input terminals (604, 605), in which a plurality of semiconductor switching devices (602, 603) are connected in series;

a plurality of output terminals (606, 607, 608) each connected to a series connected point of the plurality of series connected circuits (601, 602, 603),

wherein

the plurality of semiconductor switching devices (602, 603) are adapted to turn on and off to perform a power conversion, and

each of the plurality of semiconductor switching devices (602, 603) is the semiconductor device according to any one of claims 1 to 14.

Drawing